Blood cultures contaminated: “Bundle to achieve acceptable proportions”

Abstract

Introduction. The problem of blood cultures contamination is very frequent in hospital care, it leads to the administration of unnecessary antibiotics and prolongs hospitalization. Main goal. To apply a bundle to reduce the proportion of blood culture contamination. Secondary objectives. Carry out an anonymous survey to detect improvement opportunities in the blood culture extraction technique. Methodology. Study design: Quasi-experimental study that evaluated the proportion of blood culture contamination before and after implementing an own Bundle. The baseline proportion of blood culture contamination was determined (Jan-July 2022), the intervention was performed (August 2022) and the post-intervention contamination proportion was established (Sep-April 2023). Intervention: The structure, procedure, and staff knowledge were analyzed through an own survey to detect improvement areas. Laboratory technicians were trained on the sampling procedure through a simulation using an artificial arm. A six measurements bundle´s was designed, the blood culture taking procedure was adapted, and the staff was trained. Statistical analysis. The proportion of contaminated blood cultures between the pre and post periods was analyzed using Chi2 and the relationship between the proportion of the pre and post period vs the literature (3.00% acceptable contamination) using the Z test for a proportion. A p<0.05 was considered statistically significant. Stata 8 software was used. Results. During the study, a total of 3,965 blood cultures were analyzed. Of these, 1,978 correspond to the pre-intervention period and 1,987 correspond to the post-intervention period. During the pre-intervention, 61 blood cultures contaminated were detected (3.08% vs 3.00% in the literature, p:0.5866) while in the post-intervention stage there were 30 blood cultures contaminated (1.51% vs 3.00% in the literature, p:0.0000). The proportion of blood cultures contaminated was halved, 3.08% vs 1.51%, p: 0.001. An anonymous survey was carried out before and after the intervention, achieving improvements in the technique of taking blood cultures. Conclusion. The implementation of the own Bundle for the extraction of blood cultures, allowed to reduce the proportion of contamination by half. The analysis of the survey allowed us to identify opportunities for improvement in the blood culture sample collection technique.